Residential Rate Study for the
Kansas Corporation
Commission
Final Report
Daniel G. Hansen
Michael T. O'Sheasy
April 11, 2012
Christensen Associates Energy Consulting, LLC
800 University Bay Drive, Suite 400
Madison, WI 53705-2299
Voice 608.231.2266 Fax 608.231.2108
Table of Contents
Executive Summary .......................................................................................................... 1
1. Introduction and Purpose of the Study....................................................................... 6
2. Description of the Rate Structures Included in the Study ........................................ 7
2.1 Base residential rate .................................................................................................. 8
2.2 Flat rate ..................................................................................................................... 9
2.3 Straight fixed variable (SFV) rate ........................................................................... 10
2.4 Inclining block rate (IBR) ....................................................................................... 11
2.5 Time-of-use (TOU) rate .......................................................................................... 12
2.6 Day-type TOU rate ................................................................................................. 13
3. Rate Design Methodology .......................................................................................... 14
3.1 Prepare customer usage data ................................................................................... 14
3.2 Rate design summary .............................................................................................. 15
3.2.1 Flat rate ........................................................................................................ 15
3.2.2 Straight-fixed variable rate .......................................................................... 15
3.2.3 Inclining block rate ...................................................................................... 16
3.2.4 Time-of-use rate ........................................................................................... 17
3.2.5 Day-type TOU rate ...................................................................................... 18
4. Bill Impacts .................................................................................................................. 19
4.1 Flat rate ................................................................................................................... 19
4.2 Straight fixed variable rate ...................................................................................... 22
4.3 Inclining block rates ................................................................................................ 25
4.4 IBR and SFV ........................................................................................................... 27
4.5 Time-of-use rates .................................................................................................... 29
4.6 Day-type TOU rates ................................................................................................ 30
4.7 Summary of bill impacts ......................................................................................... 32
5. Load Response............................................................................................................. 33
5.1 SFV ......................................................................................................................... 34
5.2 IBR .......................................................................................................................... 35
5.3 TOU ........................................................................................................................ 35
5.4 Day-Type TOU ....................................................................................................... 36
6. Potential for Utility Revenue Attrition ..................................................................... 37
6.1 Revenue attrition due to customer self selection .................................................... 37
6.2 Revenue attrition due to customer demand response.............................................. 39
7. Summary and Conclusions ........................................................................................ 41
i
CA Energy Consulting
Tables
Table ES.1: Summary of Bill Impacts by Rate Structure, KCP&L
Table ES.2: Summary of Bill Impacts by Rate Structure, Westar
Table ES.3: Summary of Bill Impacts by Rate Structure, Midwest
Table 2.1: Base Residential Rates, KCP&L
Table 2.2: Base Residential Rates, Westar
Table 2.3: Base Residential Rates, Midwest
Table 3.1: Number of Customers used in the Analysis, by Utility
Table 3.2: Flat Rate ($/kWh), by Utility
Table 3.3: SFV Rates, by Utility
Table 3.4: Inclining Block Definitions and Prices ($/kWh), KCP&L
Table 3.5: Inclining Block Definitions and Prices ($/kWh), Westar
Table 3.6: Inclining Block Definitions and Prices ($/kWh), Midwest
Table 3.7: TOU Periods and Prices ($/kWh), KCP&L
Table 3.8: TOU Periods and Prices ($/kWh), Westar
Table 3.9: Day-type TOU Prices by Day Type ($/kWh), KCP&L
Table 3.10: Day-type TOU Prices by Day Type ($/kWh), Westar
Table 4.1: Share of High and Low Bill Impacts, by Utility
Table 4.2: Summary of Bill Impacts by Rate Structure, KCP&L
Table 4.3: Summary of Bill Impacts by Rate Structure, Westar
Table 4.4: Summary of Bill Impacts by Rate Structure, Midwest
Table 5.1: Percentage Changes in Usage by Season and Utility, SFV
Table 5.2: Percentage Changes in Usage by Season and Utility, IBR
Table 5.3: Percentage Changes in Usage by TOU Pricing Period and Utility
Table 5.4: Percentage Changes in Usage by Day-type TOU Day Type
Table 6.1: Revenue Attrition Due to Customer Self Selection
Table 6.2: Elasticity Assumptions by Rate and Scenario
Table 6.3: Revenue Attrition Due to Customer Demand Response, Expected Elasticity
Table 6.4: Revenue Attrition Due to Customer Demand Response, High Elasticity
Table 6.5: Revenue Attrition Due to Customer Demand Response, Low Elasticity
ii
4
4
4
8
8
9
15
15
16
16
16
17
18
18
19
19
27
33
33
33
35
35
36
37
38
39
40
40
40
CA Energy Consulting
Figures
Figure 4.1: Percentage Flat Rate Bill Impacts, KCP&L ............................................................... 20
Figure 4.2: Percentage Flat Rate Bill Impacts, Westar with Peak Management .......................... 21
Figure 4.3: Percentage Flat Rate Bill Impacts, Westar ................................................................. 21
Figure 4.4: Percentage Flat Rate Bill Impacts, Midwest ............................................................... 22
Figure 4.5: Percentage SFV Rate Bill Impacts, KCP&L .............................................................. 23
Figure 4.6: Percentage SFV Rate Bill Impacts, Westar ................................................................ 24
Figure 4.7: Percentage SFV Rate Bill Impacts, Midwest.............................................................. 24
Figure 4.8: Percentage IBR Bill Impacts, KCP&L ....................................................................... 25
Figure 4.9: Percentage IBR Bill Impacts, Westar ......................................................................... 26
Figure 4.10: Percentage IBR Bill Impacts, Midwest..................................................................... 26
Figure 4.11: Percentage IBR+SFV Bill Impacts, KCP&L ........................................................... 28
Figure 4.12: Percentage IBR+SFV Bill Impacts, Westar ............................................................. 28
Figure 4.13: Percentage IBR+SFV Bill Impacts, Midwest ........................................................... 29
Figure 4.14: Percentage TOU Bill Impacts, KCP&L ................................................................... 30
Figure 4.15: Percentage TOU Bill Impacts, Westar ..................................................................... 30
Figure 4.16: Percentage Day-Type TOU Bill Impacts, KCP&L .................................................. 31
Figure 4.17: Percentage Day-Type TOU Bill Impacts, Westar .................................................... 32
iii
CA Energy Consulting
Executive Summary
This report documents a residential rate study that Christensen Associates Energy
Consulting, LLC (CA Energy Consulting) conducted on behalf of the Kansas Corporation
Commission (KCC). The KCC is interested in studying rates that can encourage
conservation and/or provide efficient rates. "Conservation" refers to providing customers
with incentives to reduce energy consumption. "Efficient rates" are those that provide
customers with prices that reflect the marginal cost to serve them, which in theory leads to
the most efficient use of resources (e.g., electricity generators). These two goals do not
always coincide. For example, a TOU rate may have low off-peak prices to reflect the fact
that only low-cost generators are needed to serve off-peak loads. While this price is
efficient, it provides less incentive to conserve in off-peak hours than an equivalent flat
price (in which the price is the same across all hours).
We used data from Kansas City Power & Light (KCP&L), Westar Energy (Westar), and
Midwest Energy (Midwest) to analyze several alternative residential rate structures. The
rate structures included in the study are:
• Flat rate;
• Straight-fixed variable (SFV) rate;
• Inclining block rate (IBR);
• Time-of-use (TOU) rate; and
• Day-type TOU rate.
The flat rate is included primarily as a reference case, in which the price does not vary by
time or with the level of customer use. SFV rates address the utility's incentive to promote
conservation and energy efficiency by increasing the fixed monthly customer charge and
reducing the throughput volumetric rate, thereby recovering all utility fixed costs through
fixed charges rather than through volumetric rates. An IBR is intended to provide an
incentive to conserve by increasing the rate a customer pays as its usage level increases.
TOU rates are intended to provide efficient price signals by charging rates that are based on
the average cost to serve customers. TOU rates therefore give customers an incentive to
reduce usage during high-cost hours (e.g., summer afternoons) and increase usage during
low-cost hours (e.g., overnight hours). Day-type TOU rates add a "dynamic" component to
TOU rates that provides customers with a significant incentive to reduce usage on the
hottest, most costly days to serve them.
Each of these rate structures affects customers differently depending on their usage levels
and patterns. The relationship between bill impacts and customer usage levels is of interest
because stakeholders often wish to avoid adverse bill impacts for low-income customers,
and low-income customers are often believed to use less electricity than other customers.
The advantages and disadvantages of each rate structure are described in the full report.
Research Approach
The following steps were used to evaluate the alternative rate structures of interest:
1) Design revenue-neutral alternative residential rates for each utility;
2) Estimate customer-level bill impacts for each rate structure at historical loads;
1
CA Energy Consulting
3) Evaluate the relationship between bill impacts and customer usage levels;
4) Simulate the changes in customer usage levels and patterns (i.e., "demand
response") in response to the new rate structures; and
5) Estimate the potential for utility revenue loss (revenue attrition) due to mispricing
the new rate options.
Design revenue-neutral alternative residential rates for each utility: Separate revenueneutral rates were designed for each utility using utility-specific residential customer usage
data and Southwest Power Pool (SPP) price data (to design the TOU and day-type TOU
rates). The rates were designed so that they produced the same amount of total revenue as
the current rate produces.
Estimate customer-level bill impacts for each rate structure at historical loads: Each
customer's bill was calculated for both their current rate and each alternative rate structure
using historical loads.
Evaluate the relationship between bill impacts and customer usage levels: To evaluate the
relationship between bill impacts and customer usage levels, the bill impacts are displayed
as scatter plots against each customer's average monthly usage (in kWh). This allows for
an easy examination of how bill impacts vary with customer usage level.
Simulate customer demand response to each rate structure: Simulation was used to estimate
the changes in load that could be expected from each rate structure. We used evidence
from existing studies on customer price responsiveness to provide estimates of the potential
magnitude of the load changes (which, depending on the rate, could be an overall increase,
an overall reduction, or shifting from high- to low-cost hours) that might be expected from
each rate structure.
Estimate the potential for utility revenue loss (revenue attrition) due to mispricing the new
rate options: The final step was to examine the potential for utility revenue attrition, or lost
revenues, due to self selection and demand response. Revenue attrition due to customer
self selection can occur when the utility sets rates without accounting for the tendency of
customers to select the rate that is most beneficial for them (i.e., gives them the lowest bill).
Revenue attrition due to customer demand response can occur when the utility sets rates
using historical load profiles but customers modify their usage patterns in response to the
pricing signals of their new rate.
2
CA Energy Consulting
Research Implementation
We used utility-specific customer data to calculate bill impacts for each rate structure.
KCP&L and Westar provided us with 2007 hourly data from their residential load research
samples. Midwest did not have a load research sample, and instead provided us with 2009
monthly billing data for its residential customers.
The rates within the alternative structures were set to produce the same total revenue as the
existing base residential rate for the available sample customers. Therefore, the first step in
the rate design process was to calculate the total revenue (accounting for the sample
weights) from the base residential rate. The assumptions used when setting the rates were
(a) all customers are on the rate (i.e., there is no customer selection issue), and (b) the
historical load profiles are retained (i.e., we ignore the potential effect of demand response
on customers’ usage and bills).
For each of the rate structures, we calculated customer-level bills using the available
customer-level load data, the "base" residential rates, and the newly designed rates. We
then calculated "instant" bill impacts, which are the bill impacts before the customers
modify their load profiles in response to the new price signals. For ease of analysis, scatter
plots of bill impacts verses customer’s average monthly usage were used. For some of the
rate structures, such as IBR or SFV, the bill impacts are strongly related to customer size.
For others, such as TOU, this is not the case.
Research Results
Bill Impacts
Tables ES.1 through ES.3 provide results that summarize the bill impact analyses. Four
statistics are provided for each utility and rate structure:
• The share of customers that experienced a bill increase of 10% or more on the new
rate structure;
• The share of customers that experienced a bill decrease of 10% or more on the new
rate structure;
• The average percentage bill impact for customers who use an average of 500 kWh
per month or less; and
• The average percentage bill impact for customers who use an average of 2,000 kWh
per month or more.
These statistics are intended to facilitate comparisons of bill impacts across rate structures
and utilities. Following are the key observations from these tables:
• The flat, TOU, and day-type TOU rates do not produce large percentage load
impacts for very many customers (as shown in the "Greater than 10% column").
• The bill impacts for the flat, TOU, and day-type TOU rates are not strongly related
to customer usage levels (as illustrated by the similarity of the average bill impacts
in the "Low Use " and "High Use" columns).
3
CA Energy Consulting
•
•
The high customer charge in the SFV rate leads to large bill increases for low-use
customers (e.g., 27.4 percent for KCP&L's low-use customers). The percentage bill
decreases for high-use customers on this rate structure are smaller in magnitude
(e.g., 5.7 percent for KCP&L's high-use customers).
Despite the fact that IBR and SFV have opposite effects by customer usage levels,
combining the two rate structures is not enough to offset SFV's adverse bill impacts
for low-use customers.
Table ES.1: Summary of Bill Impacts by Rate Structure, KCP&L
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
TOU
Day-type TOU
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
1.3%
0.0%
15.1%
0.0%
4.9%
0.0%
3.9%
0.0%
0.3%
0.0%
0.3%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
0.1%
27.4%
-6.6%
21.2%
-0.5%
-0.5%
High Use (>2,000
kWh/mo.)
0.6%
-5.7%
10.4%
2.6%
-0.2%
-0.5%
Table ES.2: Summary of Bill Impacts by Rate Structure, Westar
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
TOU
Day-type TOU
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
0.0%
0.0%
35.9%
6.6%
5.6%
0.0%
28.8%
0.0%
0.0%
0.0%
0.0%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
-0.1%
46.6%
-1.5%
42.2%
0.1%
1.4%
High Use (>2,000
kWh/mo.)
2.6%
-10.1%
8.9%
-4.8%
1.9%
1.5%
Table ES.3: Summary of Bill Impacts by Rate Structure, Midwest
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
0.0%
0.0%
19.5%
0.4%
6.0%
0.0%
13.7%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
-2.2%
20.7%
-7.3%
16.7%
High Use (>2,000
kWh/mo.)
3.9%
-8.8%
17.9%
1.9%
The customer-level bill impacts shown above are those that occur before customers take
actions to adapt to the new rate structures (e.g., by shifting or reducing load). Of course,
the goal of most of these rate structures is to provide customers with incentives to change
behavior. The primary incentive goal of each rate structure can be summarized as follows:
4
CA Energy Consulting
•
•
•
•
SFV: Eliminates the utility's disincentive to encourage conservation and energy
efficiency. As a side effect, SFV reduces the customer-level incentive to conserve
because the volumetric rate has been reduced.
IBR: Discourages increases in consumption levels, particularly for high-use
customers who face the high tail-block price. Note that low-use customers may
experience a decrease in their incentive to conserve because they face the relatively
low initial block price.
TOU: Encourages customers to shift intra-day load from peak to off-peak hours.
Day-type TOU: Builds upon standard TOU by providing added incentives to
reduce usage on high-cost days.
Demand Response
To evaluate the potential magnitude of the usage changes described above, we developed
simple elasticity-based models to simulate the changes in usage for each of these rate
structures. The results of these simulations show that SFV leads to small increases in
overall usage; IBR leads to small decreases in overall usage; TOU leads to decreases in
peak-period usage and increases in off-peak period usage; and day-type TOU produces
larger shifts of usage from peak to off-peak periods on higher-priced days.
Revenue Attrition
Finally, the report examined the potential for utility revenue attrition (recovering less
revenue than forecast) due to customer self selection and demand response. That is, when
the utility sets the rates for an optional pricing program, it does not know which customers
will select the rate, or how the customers who select the rate will modify their load profiles
in response to the new price signals. Our analysis provided an indication of the scale of
this potential problem by assuming that customers select the rate that provides them with
the lowest bill (customer self selection); and by simulating customer demand response
using a range of price responsiveness parameters (i.e., price elasticities). The results
indicated that both types of revenue attrition (i.e., due to customer self selection and
demand response) are more pronounced for SFV and IBR than they are for TOU and daytype TOU.
5
CA Energy Consulting
1. Introduction and Purpose of the Study
This report documents a residential rate study that Christensen Associates Energy
Consulting, LLC (CA Energy Consulting) conducted on behalf of the Kansas Corporation
Commission (KCC). The KCC is interested in studying rates that can encourage
conservation and/or provide efficient rates. "Conservation" refers to providing customers
with incentives to reduce energy consumption. "Efficient rates" are those that provide
customers with prices that reflect the marginal cost to serve them, which in theory leads to
the most efficient use of resources (e.g., electricity generators). 1 These two goals do not
always coincide. For example, a TOU rate may have low off-peak prices to reflect the fact
that only low-cost generators are needed to serve off-peak loads. While this price is
efficient, it provides less incentive to conserve in off-peak hours than an equivalent flat
price (in which the price is the same across all hours).
We used data from Kansas City Power & Light (KCP&L), Westar Energy (Westar), and
Midwest Energy (Midwest) to analyze several alternative residential rate structures. The
rate structures included in the study are:
• Flat rate;
• Straight-fixed variable (SFV) rate;
• Inclining block rate (IBR);
• Time-of-use (TOU) rate; and
• Day-type TOU rate. 2
Each of these rate structures is capable of furthering progress toward encouraging
conservation or efficient pricing. The advantages and disadvantages of each are described
in the report.
Separate rates were designed for each utility using utility-specific residential customer
usage data and Southwest Power Pool (SPP) price data (to design the TOU and day-type
TOU rates).
The primary goals of the evaluation are the following:
• Design revenue-neutral alternative residential rates for each utility;
• Estimate customer-level bill impacts for each rate structure at historical loads;
• Evaluate the relationship between bill impacts and customer usage levels; and
• Estimate the potential for utility revenue loss (revenue attrition) due to mispricing
the new rate options.
1
The marginal cost of electricity in a particular hour represents the forward-looking change in the cost of
generating and delivering electric power that is caused by a change in load in that hour. With the advent of
competitive regional wholesale markets, hourly wholesale prices are generally interpreted as representing
marginal costs. Retail prices that reflect time-based changes in wholesale costs (e.g., averaged over certain
time periods) signal consumers about the cost of consuming power at those times, leading to efficient use of
resources.
2
We did not study TOU or day-type TOU rates for Midwest because they do not have a residential load
research sample. Hourly load data are required to design and evaluate these rate structures.
6
CA Energy Consulting
The relationship between bill impacts and customer usage levels is of interest because
stakeholders often wish to avoid adverse bill impacts for low income customers, and low
income customers are often believed to use less electricity than other customers. 3
A second goal of the evaluation is to simulate the demand response (i.e., changes in load)
that could be expected from each rate structure. Some of the rate structures (e.g., day-type
TOU) provide customers with the incentive to reduce load in peak periods (e.g., summer
weekday afternoon hours) and we will use existing evidence on customer price
responsiveness to provide estimates of the potential magnitude of the load reductions that
might be expected from each rate structure.
A third goal is to examine the potential for utility revenue attrition, or lost revenues, due to
customer self selection and demand response. Revenue attrition due to customer selfselection can occur when the utility sets rates without accounting for the tendency of
customers to select the rate that is most beneficial for them (i.e., gives them the lowest bill).
Revenue attrition due to customer demand response can occur when the utility sets rates
using historical load profiles and customers modify their usage patterns in response to the
pricing signals of their new rate.
After this introductory section, Section 2 describes each of the rate structures included in
the study. Section 3 describes our methodology for designing each rate structure. Section
4 presents the estimated bill impacts for each rate structure and utility. Section 5 contains
estimates of the customer load reductions and/or load shifting in response to each rate
structure. Section 6 contains an analysis of the potential for utility revenue attrition due to
customer self selection and demand response. Section 7 provides a summary and
conclusions.
2. Description of the Rate Structures Included in the Study
In the sub-sections below, we describe each of the rate structures included in the study. For
each of the new structures, we evaluate each structure according to a range of criteria:
• Economic efficiency: the extent to which prices reflect marginal costs.
• Conservation incentives: the extent to which prices encourage customers to use less
energy.
• Simplicity/transparency for customer: reflects how easily the customer can
understand the rates.
• Stability in utility revenues and customer bills: reflects the variability in revenues
and bills as changes occur in system conditions or weather.
• Utility administrative costs: how costly the rate is to administer.
• Metering requirements: the meter technology required to bill the rate.
The rating given for each of these items is qualitative in nature. That is, the exact rating
may depend on a variety of factors. The ratings given here are only intended to facilitate
comparison across rate structures.
3
We may explore the extent to which low income customers are also low use customers in Kansas, subject to
data availability and stakeholder interest.
7
CA Energy Consulting
2.1 Base residential rate
Existing utility-specific residential rates (some of which date back to 2007) were used as
the basis for all bill impact analyses in the study.
KCP&L's rates are shown in Table 2.1 below. The General Use rate shown in the second
column applies to 91 of the 95 customer load profiles that we used in the analysis.
Table 2.1: Base Residential Rates, KCP&L
Rate Component
Customer Charge ($ per customer month)
Summer Energy 1st 1,000 kWh ($/kWh)
Summer Energy over 1,000 kWh ($/kWh)
Winter Energy 1st 1,000 kWh ($/kWh)
Winter Energy over 1,000 kWh ($/kWh)
General Use Rate
$9.07
$0.08899
$0.08899
$0.08037
$0.08003
Water Heater Rate
$9.07
$0.08899
$0.08899
$0.05177
$0.07910
Westar's residential rates are shown in Table 2.2 below. In addition to the rates in the table,
10 of the 87 customers in the load research sample are on the Peak Management Rate,
which has a $10 per customer-month customer charge, flat energy charge of $0.043189 per
kWh, and seasonal demand charges of $1.65 per kW in the winter and $5.45 per kW in the
summer. Because this rate tends to produce a lower average rate paid, the bill impacts for
the Peak Management customers tend to be quite high. (There is only one rate per
alternative structure that applies to all customers.) Therefore, we typically present bill
impacts that assume that these customers are on the "standard" residential rate.
Table 2.2: Base Residential Rates, Westar
Rate Component
Customer Charge ($ per customer month)
1st 500 kWh ($/kWh)
Next 400 kWh ($/kWh)
All additional kWh ($/kWh)
Winter Rate
$8.00
$0.067892
$0.067892
$0.056045
Summer Rate
$8.00
$0.067892
$0.067892
$0.081240
Table 2.3 shows the residential rates that we used for Midwest. The have a declining block
structure in the non-summer months, though the decrease in price is not large across
blocks.
8
CA Energy Consulting
Table 2.3: Base Residential Rates, Midwest
Month
Jan-09
Feb-09
Mar-09
Apr-09
May-09
Jun-09
Jul-09
Aug-09
Sep-09
Oct-09
Nov-09
Dec-09
Customer Charge
($/cust.-mo.)
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
$ 13.00
Energy Block 1
($/kWh)
$ 0.0888
$ 0.0873
$ 0.0874
$ 0.0905
$ 0.0904
$ 0.0926
$ 0.0906
$ 0.0892
$ 0.0872
$ 0.0898
$ 0.0904
$ 0.0938
Energy Block 2
($/kWh)
$ 0.0818
$ 0.0803
$ 0.0804
$ 0.0835
$ 0.0834
$ 0.0856
n/a
n/a
n/a
$ 0.0828
$ 0.0834
$ 0.0868
Energy Block 3
($/kWh)
$ 0.0758
$ 0.0743
$ 0.0744
$ 0.0775
$ 0.0774
$ 0.0796
n/a
n/a
n/a
$ 0.0768
$ 0.0774
$ 0.0808
2.2 Flat rate
Description
A flat rate is the simplest tariff structure. For our analysis, this structure consists only of a
single price per kWh. That is,
Monthly Bill = Flat Price ($/kWh) x Monthly Usage (kWh).
More commonly, the flat rate is combined with a monthly customer charge. 4 That is,
Bill = Customer Charge + Flat Rate ($/kWh) x Monthly Usage (kWh).
The distinguishing characteristic of a flat rate is that the marginal price to the customer
does not change with the level of usage or over time.
Economic Efficiency
Rating: Low
Notes: The price does not vary with expected or actual market conditions. The price tends
to reflect average costs more than marginal costs.
Conservation Incentives
Rating: Low 5
Notes: The price does not vary with the level of usage. The price is below the cost to serve
in high-cost hours. The price exceeds the cost to serve in low-cost hours.
Simplicity / Transparency for Customer
Rating: High
Notes: It is easy for a customer to determine the change in bill associated with a change in
consumption, as the price does not change by time or with the level of usage.
4
5
The customer charge may also be expressed as dollars per day of service.
In off-peak hours, the conservation incentive could be regarded as "high" relative to the cost to serve.
9
CA Energy Consulting
Stability in Utility Revenues and Customer Bills
Rating: Medium
Notes: Revenues / bills change with usage levels, which are affected by weather, economic
conditions, etc.
Utility Administrative Costs
Rating: Low
Notes: Bill calculation is easy and rates are set infrequently.
Metering Requirements
A standard energy meter is the only requirement.
2.3 Straight fixed variable (SFV) rate
Description
SFV rates are flat rates in which all fixed costs are recovered through a monthly customer
charge. 6 This rate structure is intended to remove the utility's disincentive to promote
conservation and energy efficiency that occurs when some or all fixed costs are recovered
through volumetric rates.
Economic Efficiency
Rating: Medium
Notes: By recovering all fixed costs through the customer charge, the energy price ought to
more closely approximate the marginal cost of energy. However, the energy price does not
vary with expected or actual market conditions.
Conservation Incentives
Rating: Low for the customer.
Notes: The customer-level incentive to conserve is lower relative to a flat rate in which
fixed costs are partly recovered through the energy price. However, the utility has an
increased incentive to promote conservation, which may offset this effect.
Simplicity / Transparency for Customer
Rating: High
Notes: It is easy for a customer to determine the change in bill associated with a change in
consumption, as the price does not change by time or with the level of usage.
Stability in Utility Revenues and Customer Bills
Rating: High
Notes: Revenues / bills change with usage levels, but by much less than on a standard flat
rate. Utility revenues to recover fixed costs do not vary (on a per-customer basis).
Utility Administrative Costs
Rating: Low
Notes: Bill calculation is easy and rates are set infrequently.
6
The energy rate does not need to be flat in general, but that is how we designed it for this study.
10
CA Energy Consulting
Metering Requirements
A standard energy meter is the only requirement.
2.4 Inclining block rate (IBR)
Description
Under block rates, the per-unit price of electricity changes with the level of consumption.
Block rates may be used to achieve a variety of goals. Inclining block rates, in which the
rate increases with the level of usage, may be used to encourage conservation or subsidize
low-use customers.
We use a three-block rate, which is billed as follows:
Monthly Bill = Customer Charge + Block 1 Rate ($/kWh) x Block 1 Usage (kWh)
+ Block 2 Rate ($/kWh) x Block 2 Usage (kWh)
+ Block 3 Rate ($/kWh) x Block 3 Usage (kWh)
For an inclining block rate, the rate in the first block is lower than the rate in the second
block, which in turn is lower than the rate in the next block. Section 3 describes the
methods used to set the block sizes and rates.
Economic Efficiency
Rating: Low
Notes: The price does not vary with expected or actual market conditions. The tail-block
price is likely to exceed marginal costs in most hours of the year.
Conservation Incentives
Rating: High for high-use customers, low for low-use customers
Notes: Customers who use enough energy to reach the high-cost blocks face a high price at
the margin. Low-use customers have a conservation incentive that is lower than it would
be under an equivalent flat rate.
Simplicity / Transparency for Customer
Rating: Low to medium
Notes: Understanding how changes in usage affect bills requires the customer to understand
(and possibly forecast) its total usage levels as well as the tariff block sizes and associated
rates.
Stability in Utility Revenues and Customer Bills
Rating: Low
Notes: The high tail-block price, combined with variability in tail-block usage levels, can
produce relatively high variability in utility revenues and customer bills.
Utility Administrative Costs
Rating: Low
11
CA Energy Consulting
Notes: Bill calculation is easy and rates are set infrequently. Block rates may be more
difficult to implement than flat rates depending upon the capabilities of the billing system
and the ability to determine appropriate block thresholds and rates.
Metering Requirements
A standard energy meter is the only requirement.
2.5 Time-of-use (TOU) rate
Description
TOU rates contain prices that vary across the hours of the day. These rates are fixed within
a time-of-use period and do not respond to changing system cost conditions. The primary
motivation for TOU rates is that electricity costs vary across the hours of the day in
reasonably predictable ways. By establishing different rates for different periods of the
day, it is possible for rates to be more reflective of average differences in the cost to serve.
TOU rates provide customers with an incentive to reduce peak-period usage, perhaps by
shifting it to lower-cost hours. For this study, the TOU rates have two pricing periods
(peak and off-peak) per season (summer is defined as May through September and all other
months are defined as non-summer).
Economic Efficiency
Rating: Medium
Notes: TOU rates account for average variations in electricity costs by hour and day type.
Therefore, the rates can reflect expected marginal costs to serve by time periods. However,
on any particular day, there can still be a substantial difference between, for example, the
TOU peak price and the peak-period marginal energy costs on that day.
Conservation Incentives
Rating: Low in off-peak hours, higher in peak-hours
Notes: Relative to a flat rate, the incentive to conserve is higher during peak hours and
lower during off-peak hours. 7
Simplicity / Transparency for Customer
Rating: Medium
Notes: In order for a customer to understand how changes in usage affect their bill, the
customer must know the relevant TOU time periods and the applicable rates. However, the
schedule of rates does not change.
Stability in Utility Revenues and Customer Bills
Rating: Medium
Notes: Relative to a flat rate, TOU rates may increase the variability of utility revenues and
customer bills because of the higher peak-period prices. However, costs are expected to be
higher during those hours as well, so the variability of net revenues may not increase.
Utility Administrative Costs
7
This assumes that the flat rate and the TOU rate are set to recover the same level of revenue. Therefore,
setting the peak period rate higher than the flat rate requires the off-peak rate to be lower than the flat rate.
12
CA Energy Consulting
Rating: Low
Notes: Within each TOU pricing period, the bill calculation is simply the metered usage
multiplied by the applicable rate.
Metering Requirements
A time-of-use energy meter is required.
2.6 Day-type TOU rate
Description
A day-type TOU rate allows TOU rates to vary with expected system conditions. For
example, the rate may consist of three sets of summer TOU rates (we use a standard TOU
rate for the non-summer months):
• "Red", or high rates that apply to a maximum of 5 summer non-holiday weekdays;
• "Yellow", or medium rates that apply to a maximum of 15 summer non-holiday
days; and
• "Green", or low rates that apply to the remaining summer days.
Customers are provided with the green, yellow, and red TOU rates at the beginning of the
year or season, but do not know ahead of time which of the three sets of rates will be in
effect on a particular day until the preceding afternoon.
This rate structure is an extension of critical peak pricing (CPP), which typically has two
day types: "critical days", in which the peak-period price is very high (sometimes $1 per
kWh or more), and all other days.
Economic Efficiency
Rating: High
Notes: Rates account for variations in electricity costs by hour and day type. Day-type
TOU moves beyond standard TOU rates by better aligning the rates with market conditions
(on a day-ahead basis).
Conservation Incentives
Rating: Low in off-peak hours, higher in peak-hours
Notes: Relative to a flat rate, the incentive to conserve is higher during peak hours and
lower during off-peak hours.
Simplicity / Transparency for Customer
Rating: Low to medium
Notes: In order for a customer to understand how changes in usage affect their bill, the
customer must know the time periods during which rates apply and the applicable rates,
and be prepared to obtain information on the following day’s type, as the schedule of TOU
rates that applies on a particular day changes with day-ahead notice.
Stability in Utility Revenues and Customer Bills
Rating: Medium
13
CA Energy Consulting
Notes: Relative to a flat rate, Day-type TOU rates may increase the variability of utility
revenues and customer bills because of the higher peak-period prices. However, costs are
expected to be higher during those hours as well, so the variability of net revenues may not
increase.
Utility Administrative Costs
Rating: Medium
Notes: The utility must develop a protocol for determining the TOU day type for the
following day and communicating that information to its customers. Also, three sets of
rates must be designed.
Metering Requirements
An interval meter is required.
3. Rate Design Methodology
3.1 Prepare customer usage data
KCP&L and Westar provided us with 2007 hourly data from their residential load research
samples. Midwest does not have a load research sample, and instead provided us with
2009 monthly billing data for its residential customers. We examined the usage data to
ensure that they provided a reasonable basis for bill comparisons under the potential
alternative rate designs. In some cases for the hourly customer data, we could "clean" a
relatively small number of observations (i.e., to remove data missing because of service
outages or metering error) and retain the customer's data. In other cases, we excluded the
customer's data entirely, typically because it appeared that the customer closed its account
during the sample timeframe.
Table 3.1 provides a summary of the number of customers for whom we received data and
the number of those customers that we retained for the analysis, for each utility. For
Midwest, we used only M system, regular residential schedule customers. For KCP&L and
Westar, we used the utility-provided sample weights to ensure that each customer was
given the proper weight in the study. 8
8
For example, utilities often over-sample high-use customers in their load research samples, to ensure that
their profile is represented in the limited number interval data that can be obtained. When using the data to
calculate a class load profile, these over-sampled customers are given less weight than other customers to
ensure that the profile properly represents the average class usage pattern.
14
CA Energy Consulting
Table 3.1: Number of Customers used in the Analysis, by Utility
Utility
KCP&L
Westar
Midwest
# of Customers in
Raw Data
105
114
4,532
# of Customers Retained
for the Analysis
95
87
3,620
3.2 Rate design summary
The rates within the alternative structures are set to produce the same total revenue as the
existing base residential rate (in Section 2.1) for the available sample customers.
Therefore, the first step in the rate design process is to calculate the total revenue
(accounting for the sample weights) from the base residential rate. The assumptions used
when setting the rates are a) all customers are on the rate (i.e., there is no customer
selection issue), and b) the historical load profiles are retained (i.e., we ignore the potential
effect of demand response on customers’ usage and bills).
3.2.1 Flat rate
The flat rate is comparatively easy to set: we simply solve for the single rate that provides
the same revenue as the base residential rate at the historical usage level. We set the
customer charge at its level in the base residential rate. Table 3.2 summarizes the flat rates
that were set for each utility. 9
Table 3.2: Flat Rate ($/kWh), by Utility
Utility
KCP&L
Westar
Midwest
Flat Rate ($/kWh)
$0.08570
$0.06779
$0.08595
3.2.2 Straight-fixed variable rate
Two steps are required to set the SFV rate. First, we use cost-of-service information to
obtain the amount of fixed costs per customer, which is then converted into a monthly
customer charge. 10 Given the revenue implied by this customer charge, we then solve for
the flat energy price that produces the same amount of total revenue (based on the sampleweighted average bill from the load research sample for KCP&L and Westar) as the base
rate. Table 3.3 summarizes the SFV rates that were set for each utility.
9
The Westar rate assumes that no customers are on the Peak Management rate. If the appropriate customers
are billed on the Peak Management rate (i.e., the customer is in the load research sample and is on the Peak
Management rate), the flat rate goes down to $0.06498.
10
It would be quite easy for us to update the results if the utilities believe that a different customer charge is
required to recover all fixed costs.
15
CA Energy Consulting
Table 3.3: SFV Rates, by Utility
Utility
KCP&L
Westar
Midwest
Customer Charge
($ per cust-mo.)
$19.72
$25.00
$26.21
Flat Rate ($/kWh)
$0.07578
$0.05198
$0.06921
3.2.3 Inclining block rate
To create the Inclining Block Rates (IBR), we examined the distribution of monthly usage
amounts for residential customers. We allow for three block prices in each season. The
thresholds were established using the distribution of customer monthly usage values in
2007 for Westar and KCP&L and 2009 for Midwest. Within each season, we attempted to
set the thresholds such that approximately one-third of the customers fall into each category
(e.g., one-third of the customers have monthly usage that reaches into the second block).
The block rates were set to be revenue neutral within season, where the first block price is
10% lower than the second block price, and the third block price is 25% higher than the
second block price. While somewhat arbitrary, these relationships between block prices
reflect the goal of increasing the customer-level incentive to conserve, at least for
customers who are exposed to the higher block prices. The resulting block definitions and
revenue neutral rates for each utility are shown in Tables 3.4 through 3.6.
Table 3.4: Inclining Block Definitions and Prices ($/kWh), KCP&L
Block Description
Block Definition
Summer First
Less than 900 kWh
Summer Second
900 kWh to 1,500 kWh
Summer Third
More than 1,500 kWh
Non-summer First
Less than 700 kWh
Non-summer Second 700 kWh to 1,000 kWh
Non-summer Third
More than 1,000 kWh
Rate
$0.07934
$0.08816
$0.11020
$0.07591
$0.08435
$0.10543
Table 3.5: Inclining Block Definitions and Prices ($/kWh), Westar
Block Description
Block Definition
Summer First
Less than 1,200 kWh
Summer Second
1,200 kWh to 1,800 kWh
Summer Third
More than 1,800 kWh
Non-summer First
Less than 600 kWh
Non-summer Second 600 kWh to 1,000 kWh
Non-summer Third
More than 1,000 kWh
16
Rate
$0.06736
$0.07485
$0.09356
$0.05518
$0.06132
$0.07665
CA Energy Consulting
Table 3.6: Inclining Block Definitions and Prices ($/kWh), Midwest
Block Description
Block Definition
Summer First
Less than 800 kWh
Summer Second
800 kWh to 1,300 kWh
Summer Third
More than 1,300 kWh
Non-summer First
Less than 500 kWh
Non-summer Second 500 kWh to 800 kWh
Non-summer Third
More than 800 kWh
Rate
$0.08211
$0.09123
$0.11404
$0.07668
$0.08520
$0.10649
An alternative, and possibly more effective, method for defining block sizes is to make
them customer specific, based on each customer’s historical usage levels. This method is
currently being tested in a residential pilot program at Commonwealth Edison. While it is
administratively more complicated, this method has the potential to provide every customer
with an increased incentive to reduce usage (relative to current or flat rates), while
maintaining revenue neutrality for all customers at their historical usage pattern. Under
“standard” inclining block pricing (in which everyone has the same block definitions), lowuse customers are likely to experience reductions in their incentive to conserve and their
bills decrease. In contrast, high-use customers are likely to experience bill increases, along
with greater incentives to conserve. We do not explicitly analyze IBR with customerspecific blocks in this study, but we may examine them in greater depth if there is sufficient
interest from the stakeholders.
3.2.4 Time-of-use rate
We set TOU rates for KCP&L and Westar. The lack of hourly load data at Midwest left us
unable to examine time-differentiated rates for their residential customers.
Under the assumption that TOU rates should reflect expected differences in marginal costs
by time period, and that wholesale market prices signal those marginal costs, we used 2007
hourly data on Southwest Power Pool (SPP) prices in the design of TOU rates. We
combined that data with the load research sample data to determine the TOU seasons,
pricing periods, and price ratios across pricing periods. The goal is to create pricing
periods that contain hours that are most alike in terms of marginal costs (e.g., hours of high
costs and low costs), and therefore produce peak to off-peak price ratios that reflect the
greatest difference between costs by time period. We set the summer season to be May
through September and the non-summer season to be all other months. During the summer
season, the peak hours are from 11:00 a.m. to 7:00 p.m. During the non-summer season,
the peak hours are from 6 a.m. to 10 p.m. Weekends and holidays have the same TOU
pricing periods as non-holiday weekdays. Note that these are broader peak periods than we
prefer to select, particularly in the non-summer months. While the periods match the
patterns of the SPP data, broad peak periods have the disadvantage of reducing the
customers’ ability to reduce peak load and/or shift load to off-peak hours.
As with the other rates, we set the TOU rates to be revenue neutral to the base rate (prior to
any demand response). To do this, we assumed that the price ratios across TOU pricing
periods equal the ratio of SPP prices across pricing periods (using 2007 SPP data). We
17
CA Energy Consulting
then solved for the set of rates (given the ratios) that obtains revenue neutrality. 11 Tables
3.7 and 3.8 show the resulting TOU rates for each utility.
Table 3.7: TOU Periods and Prices ($/kWh), KCP&L
TOU Pricing Period
Summer Peak
Summer Off-peak
Non-summer Peak
Non-summer Off-peak
Hours
11 a.m. to 7 p.m.
7 p.m. to 11 a.m.
6 a.m. to 10 p.m.
10 p.m. to 6 a.m.
Rate
$0.11135
$0.07134
$0.09362
$0.05849
Table 3.8: TOU Periods and Prices ($/kWh), Westar
TOU Pricing Period
Summer Peak
Summer Off-peak
Non-summer Peak
Non-summer Off-peak
Hours
11 a.m. to 7 p.m.
7 p.m. to 11 a.m.
6 a.m. to 10 p.m.
10 p.m. to 6 a.m.
Rate
$0.08777
$0.05648
$0.07422
$0.04617
3.2.5 Day-type TOU rate
Day-type TOU rates allow prices to vary with day-ahead notice during the summer months,
and therefore better reflect wholesale costs, particularly on the relatively few high-load and
high-cost days during the summer. The non-summer TOU rates and all peak and off-peak
hour definitions are identical to the values presented in Section 3.2.4. Three sets of TOU
rates are set:
• "Red", or high rates, which apply on up to five summer days;
• "Yellow", or moderate rates, which apply on up to fifteen summer days; and
• "Green", or low rates, which apply on all of the remaining summer days.
The day types were set using SPP prices, where the five highest-price (defined as the
average peak-period price) weekdays are defined as "Red", the next fifteen highest-price
weekdays are defined as "Yellow", and the remaining days are defined as "Green".
The 2007 SPP prices that we used are not high enough (even on the highest-price days) to
create a significant amount of differentiation between the prices on the different day types,
as might be seen in some years. Therefore, to illustrate a price structure more reflective of
a wider distribution of wholesale prices, we have set the Red and Yellow prices to
relatively high levels and high ratios of peak to off-peak prices (compared to 2007 values),
and solved for the Green prices that obtain revenue neutrality to the base residential rates.
The Green prices are set using the peak to off-peak price ratio observed in the SPP prices
for the Green days. Tables 3.9 and 3.10 contain the summer day-type TOU prices for
KCP&L and Westar, respectively. 12
11
This is not the only method that can be used to create revenue neutral TOU rates. For example, the peak
rate could be set at the expected market marginal cost with the off-peak rate set to obtain revenue neutrality.
12
Utilities in some states have developed rates similar to day-type TOU (e.g., critical-peak pricing, or CPP) in
which peak prices on “critical” days include an allocation of avoided capacity costs as well as energy costs,
thus resulting in substantially higher critical-day peak prices, on the order of $1.00/kWh or more.
18
CA Energy Consulting
Table 3.9: Day-type TOU Prices by Day Type ($/kWh), KCP&L
Day Type
Peak
Off-peak
Red (5 days)
$0.24000 $0.08000
Yellow (15 days)
$0.15000 $0.07500
Green (all other days) $0.10284 $0.06745
Table 3.10: Day-type TOU Prices by Day Type ($/kWh), Westar
Day Type
Peak
Off-peak
Red (5 days)
$0.24000 $0.08000
Yellow (15 days)
$0.15000 $0.07500
Green (all other days) $0.07538 $0.04967
4. Bill Impacts
For each of the rate structures, we calculated customer-level bills using the available
customer-level load data, the "base" residential rates, and each of the rates described in
Section 3. We then calculated bill impacts at the customers’ historical usage patterns,
before accounting for any possible modification in customers’ load profiles in response to
the new price signals.
The bill impacts are displayed as scatter plots against each customer's average monthly
usage (in kWh). This allows for an easy examination of both the range of magnitude of the
bill impacts, as well as how the bill impacts vary with customer usage. For some of the
rate structures, such as IBR or SFV, the bill impacts are strongly related to customer usage.
For others, such as TOU, this is not the case, as bill impacts are related to differences in
percentage of usage in peak periods.
The relationship between bill impacts and customer usage may be of interest because
customer usage is often equated with customer income levels, where smaller customers are
believed to have lower income levels. Therefore, bill impacts that adversely affect low-use
customers are believed to reflect adverse outcomes for low-income customers. Utilities
typically do not have customer income data, so it is not a straightforward exercise to
determine whether this relationship between income and usage actually exists. However,
some studies have attempted to make use of Census data to explore the link between usage
and income. One example is a recent article that examined the redistribution of income
that occurs under inclining block rates. 13 We could consider implementing a similar
analysis as an extension of this work if there is sufficient interest from the stakeholders.
The sub-sections below present bill impacts for each utility's residential customers, ordered
by rate structure.
4.1 Flat rate
Figures 4.1 through 4.4 show the customer-level bill impacts for the flat rates shown in
Table 3.2. For KCP&L (shown in Figure 4.1), there are four customers who have
13
Borenstein, Severin. "The Redistributional Impact of Non-Linear Electricity Pricing", Energy Institute at
Haas Working Paper Series, March 2010.
19
CA Energy Consulting
significantly higher bill impacts than the other customers (i.e., approximately 10 percent,
where the bill impact for majority of customers is between +/- 2 percent). These customers
are on the space heating rate, which offers lower winter rates than the standard residential
rate. Because one flat rate is set to apply to all of the residential customers, the space
heating customers experience a more adverse bill impact than the others.
Figure 4.1: Percentage Flat Rate Bill Impacts, KCP&L
12.0%
10.0%
% Flat Rate Bill Impact
8.0%
6.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-2.0%
-4.0%
Avg Monthly kWh
A similar situation is present for Westar, shown in Figure 4.2. In this case, ten of the load
research sample customers are on the Peak Management Rate, which contains a single
energy price and a demand charge. The average price on this rate is significantly below the
average price on the standard residential rate, so these customers experience a large bill
increase when they are migrated to the flat rate. Because these customers are such outliers
relative to the standard residential customers, for the remainder of the analysis we treat
them as standard rate customers, and calculate their base bills at the standard rate. This is
shown in Figure 4.3.
20
CA Energy Consulting
Figure 4.2: Percentage Flat Rate Bill Impacts, Westar with Peak Management
35.0%
30.0%
25.0%
15.0%
10.0%
5.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-5.0%
-10.0%
-15.0%
Average Monthly kWh
Figure 4.3: Percentage Flat Rate Bill Impacts, Westar
12.0%
10.0%
8.0%
6.0%
% Flat Rate Bill Impact
% Flat Rate Bill Impact
20.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-2.0%
-4.0%
-6.0%
-8.0%
Average Monthly kWh
21
CA Energy Consulting
Figure 4.4 shows the bill impacts for Midwest's residential customers. Notice that data for
many more customers are available (because we use billing data and not hourly data),
which may provide a more complete picture of the distribution of bill impacts.
Figure 4.4: Percentage Flat Rate Bill Impacts, Midwest
10.0%
8.0%
% Flat Rate Bill Impact
6.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
-2.0%
-4.0%
Average Monthly kWh
For Midwest, the bill impacts are strongly related to customer size, with bill decreases for
small customers and bill increases for large customers. This is because Midwest's base
residential rate has a declining block structure in nine of the twelve months of the year, and
the tail block rate is 1.3 cents/kWh lower than the rate in the first block. This effect is not
as pronounced for KCP&L and Westar. For KCP&L most of the customers are on a rate
that has seasonal differentiation with a declining block in the winter months, but the
magnitude of the decline is trivial ($0.00034 per kWh). For Westar, the effect of the
declining block rates in the winter months is offset by the use of inclining block rates in the
summer months.
4.2 Straight fixed variable rate
Figures 4.5 through 4.7 show the bill impacts for straight fixed variables rates in Table 3.3.
These have the same structure as the flat rates presented in Section 4.1, but with higher
customer charges set to cover all fixed costs. The basic story is the same for all three
utilities: because of the increase in the customer charge, low-use customers experience bill
increases and high-use customers experience bill decreases. The magnitude of the bill
increases for low-use customers varies somewhat across utilities. One way to quantify this
22
CA Energy Consulting
is the share of customers who experience a bill increase of more than 20 percent, which is
20 percent for Westar, 9.4 percent for Midwest, and 3.2 percent for KCP&L. 14
While Westar has the highest share of customers who experience an especially adverse bill
impact, it also has the highest share of customers who experience significant bill reductions
on SFV. Approximately 6 percent of Westar's customers experience more than a 10
percent bill decrease on SFV. In contrast, none of KCP&L's customers and only 0.4
percent of Midwest's customers achieve that level of bill reduction on SFV. 15
Figure 4.5: Percentage SFV Rate Bill Impacts, KCP&L
60.0%
50.0%
40.0%
% SFV Bill Impact
30.0%
20.0%
10.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-10.0%
-20.0%
Average Monthly kWh
14
The shares are calculated using the customer sample weights, as opposed to simply using the share of
customers in the available sample.
15
We are aware of a few utilities that have attempted to reduce the adverse effect of higher customer charges
on low-use customers by instituting a graduated customer charge, on the theory that the amount of fixed costs
that customers cause the utility to incur are related to their usage level.
23
CA Energy Consulting
Figure 4.6: Percentage SFV Rate Bill Impacts, Westar
100.0%
80.0%
% SFV Bill Impact
60.0%
40.0%
20.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-20.0%
-40.0%
Average Monthly kWh
Figure 4.7: Percentage SFV Rate Bill Impacts, Midwest
100.0%
80.0%
% SFV Bill Impact
60.0%
40.0%
20.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
-20.0%
Average Monthly kWh
24
CA Energy Consulting
4.3 Inclining block rates
The bill impacts associated with the inclining block rates introduced in Tables 3.4 through
3.6 are shown in Figures 4.8 through 4.10. IBR, in which the rate increases with the usage
level, produces bill impacts that benefit low-use customers at the expense of high-use
customers. This is the opposite of the effect of SFV rates.
Figure 4.8: Percentage IBR Bill Impacts, KCP&L
25.0%
20.0%
% IBR Bill Impact
15.0%
10.0%
5.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-5.0%
-10.0%
Average Monthly kWh
25
CA Energy Consulting
Figure 4.9: Percentage IBR Bill Impacts, Westar
20.0%
15.0%
% IBR Bill Impact
10.0%
5.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-5.0%
-10.0%
Average Monthly kWh
Figure 4.10: Percentage IBR Bill Impacts, Midwest
35.0%
30.0%
25.0%
% IBR Bill Impact
20.0%
15.0%
10.0%
5.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
-5.0%
-10.0%
-15.0%
Average Monthly kWh
The distribution of the percentage bill impacts is similar for the three utilities. Table 4.1
shows the share of customers on the high and low end of the bill impacts. Across all three
26
CA Energy Consulting
utilities, approximately five percent of the customers experience a bill increase of at least
ten percent on IBR (these are the large customers), while 40 to 50 percent of the customers
experience at least a five percent decrease (these are low-use customers).
Table 4.1: Share of High and Low Bill Impacts, by Utility
Utility
Share with 10% or higher Share with -5% or lower
KCP&L
4.9%
45.8%
Westar
5.6%
42.1%
Midwest
6.0%
50.0%
The "break-even" usage level (where the bill impact is zero) is approximately 1,500 kWh
per month for KCP&L and Westar and 1,000 kWh per month for Midwest.
4.4 IBR and SFV
It is useful to note that the alternative rate structures are not mutually exclusive. For
example, we can combine the SFV and IBR structures by simply increasing the customer
charge on IBR and re-calculating the rates to obtain revenue neutrality. This is an
intuitively appealing combination because of the potential for offsetting bill impacts. For
example, SFV tends to increase bills for low-use customers while IBR tends to reduce
them.
Figures 4.11 through 4.13 show the resulting bill impacts of a combined SFV/IBR rate.
The results indicate that the SFV bill impacts "dominate" the IBR bill impacts for low-use
customers. That is, for low-use customers, the higher customer charge produces larger
effects than the reduction in the initial block price.
Interestingly, for KCP&L and Midwest, the combination of SFV and IBR also produces
adverse bill impacts for the largest customers. The "middle class" of customers, with usage
ranging from approximately 750 to 1,500 kWh per month, tends to benefit from this
combination of rate structures.
27
CA Energy Consulting
Figure 4.11: Percentage IBR+SFV Bill Impacts, KCP&L
50.0%
40.0%
% IBR + SFV Bill Impact
30.0%
20.0%
10.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-10.0%
Average Monthly kWh
Figure 4.12: Percentage IBR+SFV Bill Impacts, Westar
100.0%
80.0%
% IBR + SFV Bill Impact
60.0%
40.0%
20.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-20.0%
-40.0%
Average Monthly kWh
28
CA Energy Consulting
Figure 4.13: Percentage IBR+SFV Bill Impacts, Midwest
100.0%
80.0%
% SFV + IBR Bill Impact
60.0%
40.0%
20.0%
0.0%
0
500
1,000
1,500
2,500
2,000
3,000
3,500
4,000
4,500
5,000
-20.0%
Average Monthly kWh
4.5 Time-of-use rates
The bill impacts associated with the TOU rates in Tables 3.7 and 3.8 are shown in Figures
4.14 and 4.15. Note that we cannot analyze TOU rates for Midwest because they do not
have the hourly usage data required to bill the rate (i.e., to obtain sales by pricing period).
The bill impacts associated with TOU rates are related to the timing of a customer's usage
(e.g., the share of usage that is in peak hours) rather than the amount of the customer's
usage, as is the case for SFV and IBR. The distributions of bill impacts in the figures
reflect this difference, showing no strong relationship between bill impacts and customer
usage levels.
The magnitude of the bill impacts for TOU rates is lower than we observed for SFV and
IBR. The vast majority of the bill impacts are within +/- 5 percent, with 98 percent of
KCP&L's customers and 90 percent of Westar's customers falling within that range.
29
CA Energy Consulting
Figure 4.14: Percentage TOU Bill Impacts, KCP&L
16.0%
14.0%
12.0%
10.0%
% TOU Bill Impact
8.0%
6.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-2.0%
-4.0%
-6.0%
Average Monthly kWh
Figure 4.15: Percentage TOU Bill Impacts, Westar
8.0%
6.0%
4.0%
% TOU Bill Impact
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-2.0%
-4.0%
-6.0%
-8.0%
Average Monthly kWh
4.6 Day-type TOU rates
The bill impacts associated with the day-type TOU rates in Tables 3.9 and 3.10 are shown
in Figures 4.16 and 4.17. As was the case with TOU rates, we are not able to analyze this
rate structure for Midwest.
30
CA Energy Consulting
This rate structure builds upon the TOU rates to provide higher price signals on days in
which the wholesale market prices are the highest. Therefore, a customer's bill impact
(prior to demand response) will depend upon the level of usage on these high-cost days.
Because wholesale market prices are driven by electricity demand, and air conditioning
load is often a significant driver of demand, it is reasonable to suppose that the customers
whose load is most weather dependent will have the most adverse bill impact on day-type
TOU rates (because they tend to use the most on the hottest, highest cost days).
The figures show that, as with TOU rates, the day-type TOU bill impacts are not strongly
related to customer size. In fact, the bill impacts are quite similar to the TOU bill impacts.
The correlation between the TOU and day-type TOU bill impacts is 0.98 for KCP&L and
0.92 for Westar. This indicates that, for the most part, a customer who is helped (or
harmed) by TOU rates will also be helped (or harmed) by day-type TOU rates.
Figure 4.16: Percentage Day-Type TOU Bill Impacts, KCP&L
16.0%
14.0%
12.0%
% Day-Type TOU Bill Impact
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
-2.0%
-4.0%
-6.0%
Average Monthly kWh
31
CA Energy Consulting
Figure 4.17: Percentage Day-Type TOU Bill Impacts, Westar
12.0%
10.0%
8.0%
% Day-Type TOU Bill Impact
6.0%
4.0%
2.0%
0.0%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
-2.0%
-4.0%
-6.0%
-8.0%
Average Monthly kWh
4.7 Summary of bill impacts
Tables 4.2 through 4.4 provide information that summarizes the figures presented in the
previous sub-sections. Four statistics are provided for each utility and rate structure:
• The share of customers that experienced a bill increase of 10% or more on the new
rate structure;
• The share of customers that experienced a bill decrease of 10% or more on the new
rate structure;
• The average percentage bill impact for customers who use an average of 500 kWh
per month or less; and
• The average percentage bill impact for customers who use an average of 2,000 kWh
per month or more.
These statistics are intended to facilitate comparisons of bill impacts across rate structures
and utilities. Following are the key observations from these tables:
• The flat, TOU, and day-type TOU rates do not produce large percentage load
impacts for very many customers (as shown in the "Greater than 10% column").
• The bill impacts for the flat, TOU, and day-type TOU rates are not strongly related
to customer usage levels (as illustrated by the similarity of the average bill impacts
in the "Low Use " and "High Use" columns).
• The high customer charge in the SFV rate leads to large bill increases for low-use
customers (e.g., 27.4 percent for KCP&L's low-use customers). The percentage bill
decreases for high-use customers on this rate structure are smaller in magnitude
(e.g., 5.7 percent for KCP&L's high-use customers).
32
CA Energy Consulting
•
Despite the fact that IBR and SFV have opposite effects by customer usage levels,
combining the two rate structures is not enough to offset SFV's adverse bill impacts
for low-use customers.
Table 4.2: Summary of Bill Impacts by Rate Structure, KCP&L
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
TOU
Day-type TOU
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
1.3%
0.0%
15.1%
0.0%
4.9%
0.0%
3.9%
0.0%
0.3%
0.0%
0.3%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
0.1%
27.4%
-6.6%
21.2%
-0.5%
-0.5%
High Use (>2,000
kWh/mo.)
0.6%
-5.7%
10.4%
2.6%
-0.2%
-0.5%
Table 4.3: Summary of Bill Impacts by Rate Structure, Westar
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
TOU
Day-type TOU
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
0.0%
0.0%
35.9%
6.6%
5.6%
0.0%
28.8%
0.0%
0.0%
0.0%
0.0%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
-0.1%
46.6%
-1.5%
42.2%
0.1%
1.4%
High Use (>2,000
kWh/mo.)
2.6%
-10.1%
8.9%
-4.8%
1.9%
1.5%
Table 4.4: Summary of Bill Impacts by Rate Structure, Midwest
Rate Structure
Flat rate
SFV
IBR
IBR + SFV
Share of Customers by Bill
Impact Amount
Greater than
Less than
10%
-10%
0.0%
0.0%
19.5%
0.4%
6.0%
0.0%
13.7%
0.0%
Average Bill Impact by Customer Usage
Low Use (<500
kWh/mo.)
-2.2%
20.7%
-7.3%
16.7%
High Use (>2,000
kWh/mo.)
3.9%
-8.8%
17.9%
1.9%
5. Load Response
The previous section examined the customer-level bill impacts that occur before customers
take actions to adapt to the new rate structures (e.g., by shifting or reducing load). Of
course, the goal of most of these rate structures is to provide customers with incentives to
change their behavior. The primary incentive goal of each rate structure can be
summarized as follows: 16
16
The flat rate is provided for comparison purposes and not because it has especially appealing incentives.
33
CA Energy Consulting
•
•
•
•
SFV: Eliminates the utility's disincentive to encourage conservation and energy
efficiency. As a side effect, SFV reduces the customer-level incentive to conserve
because the volumetric rate has been reduced.
IBR: Discourages increases in consumption levels, particularly for high-use
customers who face the high tail-block price. Note that low-use customers may
experience a decrease in their incentive to conserve because they face the relatively
low initial block price.
TOU: Encourages customers to shift intra-day load from peak to off-peak hours.
Day-type TOU: Builds upon standard TOU by providing added incentives to
reduce usage on high-cost days.
In the sub-sections that follow, we present approximate load impacts (or demand response)
that might be expected under each of the rate structures.
5.1 SFV
When a utility makes a transition from standard rates to SFV rates, the customer charge is
increased and the energy price is decreased. This reduction in the energy price reduces the
return that customers get from investing in energy efficiency (e.g., a saved kWh used to
reduce their bill by 8 cents, but under SFV it is only reduced by 6 cents). It also reduces
the incremental cost associated with increasing usage (e.g., by reducing the thermostat
setting in summer).
We simulated the expected effects of this incentive change using a simple demand model:
% change in usage = εd * % change in marginal rate
In this model, εd is the price elasticity of demand, which we have assumed to be -0.20.
This value is consistent with values that have been estimated in the literature. 17 We
determined the marginal rate for each customer on the base rate. This was done separately
for each season using the average monthly usage across the months and taking into account
the existing block rates, if applicable. The "base" marginal rate was then compared to the
SFV rate to obtain the percentage change in the marginal rate. The percentage change in
usage was then obtained by multiplying the percentage change in the marginal rate by the
elasticity of demand. The results are shown in Table 5.1.
17
For example, a RAND study from 2005 titled "Regional Differences in the Price-Elasticity of Demand for
Energy" by Bernstein and Griffin estimated long-run and short-run elasticities of electricity demand for
residential customers by region. For the West North Central region (which includes Minnesota, Iowa,
Missouri, North Dakota, South Dakota, Nebraska, and Kansas) the long-run price elasticity was -0.244 and
the short-run price elasticity was -0.163.
34
CA Energy Consulting
Table 5.1: Percentage Changes in Usage by Season and Utility, SFV
Utility
KCP&L
Westar
Midwest
Summer Winter
+3.0%
+6.8%
+4.5%
+1.1%
+2.5%
+2.6%
In summer, increases in usage induced by the lower SFV rates range from 3.0 to 6.8
percent. The increases are smaller in winter, ranging from 1.1 to 2.6 percent. These load
changes do not include effects associated with any increases in conservation activities that
may be induced by the change in utility incentives. That is, SFV removes the utility's
disincentive to promote energy and conservation, so more conservation may occur due to
increased utility involvement in demand-side management activities. However,
accomplishing this incentive change through SFV has the side effect of reducing customerlevel incentives to conserve.
5.2 IBR
The effect of IBR on customer usage levels was analyzed using the same method described
above for SFV rates. In this case, we also determined the marginal IBR block rate for each
customer and season. In general, small customers (who face a low marginal rate) have a
reduced incentive to conserve under IBR, while large customers (who face a high marginal
rate) have an increased incentive to conserve. Table 5.2 shows the results of our load
response simulations (again assuming a -0.20 elasticity of demand).
Table 5.2: Percentage Changes in Usage by Season and Utility, IBR
Utility
KCP&L
Westar
Midwest
Summer Winter
-2.3%
-0.3%
-2.8%
-3.4%
-3.7%
-3.9%
The results indicate larger reductions in winter than summer months (these results could be
reversed by modifying the rate designs), with winter usage reductions ranging from 3.4 to
3.9 percent; and summer usage reductions ranging from 0.3 to 2.8 percent.
5.3 TOU
The model used to simulate load response to TOU and Day-type TOU rates was different
from the model used to simulate load response to SFV and IBR. In this case, we focus on
the customer's incentive to shift load from peak to off-peak hours. The magnitude of the
load shifting is described by the elasticity of substitution (εs), which we assume to be 0.10.
The load response is modeled as follows:
QRTOU = exp{ln(QRbase) – εs x ln(PRTOU)}
35
CA Energy Consulting
In this equation, QRTOU is the ratio of peak to off-peak usage on the TOU rate, QRbase is the
ratio of peak to off-peak usage on the base rate, εs is the elasticity of substitution, and
PRTOU is the ratio of peak to off-peak prices on the TOU rate. 18 Table 5.3 shows the results
of the simulations for each season and utility.
Table 5.3: Percentage Changes in Usage by TOU Pricing Period and Utility
Season
Summer
Winter
TOU Period KCP&L Westar
Peak
Off-peak
Peak
Off-peak
-2.6%
+1.8%
-1.4%
+3.4%
-2.5%
+1.9%
-1.4%
+3.4%
Notice that across all seasons and both utilities, the TOU rates produce a decrease in peakperiod usage and an increase in off-peak usage. Summer peak usage declines by
approximately 2.5 percent, while off-peak usage increases by about 1.9 percent. In winter,
the reduction in peak usage is smaller, at about 1.4 percent. Off-peak usage goes up by
about 3.4 percent.
These results illustrate how, in theory, TOU rates can lead to a more efficient use of
resources (by shifting usage from higher-cost peak to off-peak hours), but not necessarily
to an overall reduction in usage. However, in practice there may be reductions in peak
usage that are not shifted to off-peak hours. For example, a customer who turns off lights
during peak hours is not likely to want to compensate by turning on more lights during offpeak hours.
5.4 Day-Type TOU
As with "standard" TOU rates, day-type TOU rates are intended to produce reductions in
peak-period usage, with the distinction that day-type TOU rates charge higher rates on days
with a higher cost to serve. In theory, this produces higher levels of demand response
during the times of greatest need and highest avoided cost.
We apply the same model and assumptions used to simulate demand response to standard
TOU rates to the day-type TOU rates. Because the winter rates do not differ between the
two types of TOU rates, we do not summarize the winter demand response again in this
section. Table 5.4 presents the simulated demand response by day type and utility for the
summer months.
18
The ratio of peak to off-peak prices on the base rate is always 1.0, so the base rates drop out of the
equation.
36
CA Energy Consulting
Table 5.4: Percentage Changes in Usage by Day-type TOU Day Type
Utility
KCP&L
Westar
TOU Period Green Yellow
Peak
Off-peak
Peak
Off-peak
-2.5%
+1.7%
-2.4%
+1.8%
-4.1%
+2.8%
-4.0%
+2.9%
Red
-6.5%
+4.4%
-6.3%
+4.5%
As expected, more demand response occurs on the high-price days than on the low-price
days. For example, for KCP&L we simulate a 6.5 percent reduction in peak-period sales
on "Red" days, but only a 2.5 percent reduction in peak-period sales on "Green" days. The
increases in off-peak usage are also correspondingly higher on the high-cost days, because
the model assumes that peak-period load reductions are shifted to off-peak hours.
6. Potential for Utility Revenue Attrition
When a utility introduces one or more optional rates, it cannot be certain which customers
will choose to participate in each rate, or how they will modify their usage patterns in
response to the price signals offered by the chosen rate. The absence of experience with
the new rate options creates two primary sources of error when setting the rates: 1)
uncertainty regarding the participants in each rate option; and 2) uncertainty regarding the
load profile of the participating customers.
When the utility sets rates, a failure to account for the fact that customers will tend to select
the rate that is most beneficial to them ("customer self selection") and then respond to the
new rate by modifying their usage patterns ("customer demand response") can lead to the
recovery of less revenue than expected, or revenue attrition. In this section, we examine
the potential level of utility lost revenues due to each of these factors.
6.1 Revenue attrition due to customer self selection
When customers are offered the choice between rates, their rate selection may be
influenced by any number of factors, including:
• Whether they are an instant winner or loser: customers may benefit from a
particular rate structure because of their current usage pattern. For example, a
customer may be able to experience an immediate bill reduction by switching from
a flat rate to a TOU rate if they use relatively little energy during peak hours.
• Price responsiveness: customers who are able to shift usage across time periods or
reduce usage on short notice may be more willing to select a rate structure in which
the price varies across hours.
• Risk aversion: some customers may be more willing than others to be exposed to
changing energy prices.
• "Hassle" costs: customers may not want to take the time to understand more
complex rates. For example, customer response to a TOU rate requires that
customers be aware of the TOU pricing periods, whereas some customers may
simply want to pay the same rate all the time, even if doing so requires paying a
premium.
37
CA Energy Consulting
Because the utility does not have full knowledge of the factors that influence customer
choice, it does not have the ability to perfectly predict which customers will select each
rate. The assumptions the utility uses when designing a rate may be violated as customers
select rates, because the customer types and aggregate load profile served by a rate may
differ from the assumed values. This can produce revenue attrition, which is a loss of
utility revenues due to customer rate choices. The lost revenue persists until the next rate
case, at which time the utility can price each rate correctly based on the actual, instead of
expected, participants and loads.
Our analysis attempts to provide an upper bound of the revenue attrition that KCP&L,
Westar, and Midwest Energy may experience due to customer self selection when SFV,
IBR, TOU and day-type TOU rates are each introduced as an optional rate. The
assumptions we use to price each rate are as follows:
• All applicable customers adopt the rate; and
• Customers do not engage in demand response (i.e., historical loads are used).
Customer choice is simulated assuming that each customer selects the rate option that
provides them with the lowest bill.
In practice, the utility can improve upon the two pricing assumptions by basing rates on its
expectation of customer enrollments and demand response (provided that the assumptions
and modeling are accepted by the Commission). In addition, the customer choice
assumption is extreme because customers may not select the rate with the lowest bill. For
example, customers who may save on a TOU rate on average may nevertheless select a flat
rate because they do not want to keep track of the TOU pricing periods. For these reasons,
the results presented here provide an overestimate of revenue attrition due to customer self
selection. However, we believe that the results are instructive regarding the potential scale
of the issue from a utility perspective.
Table 6.1 shows the revenue attrition as a percentage of revenues from the current rates, by
utility and rate structure. The results represent the percentage revenue attrition when each
rate is introduced as the sole alternative to the current rate.
Table 6.1: Revenue Attrition Due to Customer Self Selection
Rate Structure
Utility
Midwest
KCP&L
Westar
SFV
IBR
TOU
-2.9%
-2.3%
-5.1%
-3.1%
-2.9%
-3.0%
n/a
-0.6%
-1.2%
Day-Type
TOU
n/a
-0.8%
-1.3%
The magnitude of the revenue attrition is modest for the TOU and day-type TOU rates,
with utility losses ranging from 0.6 percent to 1.3 percent of current revenue. Revenue
attrition is considerably higher for the SFV and IBR rates, with losses ranging from 2.3
percent to 5.1 percent of current revenue.
38
CA Energy Consulting
These results are consistent with the bill impacts shown in Section 4 in that larger revenue
attrition is observed for rates with more dispersed bill impacts (i.e., SFV and IBR). For
example, SFV tends to harm low-use customers. Therefore, when given the choice
between their current rate and SFV, low-use customers will tend to remain on their current
rate. However, the high-use customers that benefit from SFV will leave the current rate,
leading the utility to lose revenue from those customers. The findings therefore indicate
the importance of accounting for customer self selection when pricing SFV and TOU rates.
6.2 Revenue attrition due to customer demand response
In this section, we examine the potential for utility revenue attrition due to customer
demand response. That is, if the utility sets its rates using historical loads instead of loads
that anticipate the load response of the customers, the utility may lose revenues as
customers shift usage from high-priced to low-priced periods. For this analysis, we assume
that all customers are on the new rate (e.g., TOU or IBR) and examine the utility revenues
lost as customers modify their usage level and pattern in response to the new rate. The
demand response model applied for each rate structure is described in Section 5.
Because the results depend upon the customers' level of price responsiveness, we simulated
outcomes using a range of elasticity assumptions. Table 6.2 shows the elasticities used for
the "expected" scenario; the "high", or very price responsive scenario; and the "low", or not
very price responsive scenario. The elasticities of demand were derived from the RAND
study described in footnote 18. The elasticities of substitution were based on results from
the California Statewide Pricing Pilot, which examined customer load shifting due to TOU
and critical peak pricing rate programs. 19
Table 6.2: Elasticity Assumptions by Rate and Scenario
Scenario
Expected
High
Low
Elasticity of Demand Elasticity of Substitution
(SFV and IBR)
(TOU and Day-Type TOU)
-0.20
0.10
-0.40
0.15
-0.10
0.05
Tables 6.3 through 6.5 show the results for the expected, high, and low elasticity scenarios,
respectively. Notice that SFV rates actually lead to the utility recovering more revenue
than it did under current rates. Because SFV reduces energy rates, it costs customers less
to increase their usage than it did under current rates. They respond to this reduction in the
marginal price by increasing their usage level, which, in turn, leads to an increase in utility
revenues.
IBR leads to a reduction in utility revenue as high-use customers reduce usage in response
to the high tail-block price signal. Note that, because the first block price is lower than the
current rate, IBR gives low-use customers an incentive to increase usage. However, the
19
"Impact Evaluation of the California Statewide Pricing Pilot", Charles River Associates, 2005, pages 91
and 97.
39
CA Energy Consulting
effect for low-use customers is outweighed by the effect for high-use customers, leading to
an overall reduction in utility revenues.
TOU and day-type TOU rates lead to small reductions in utility revenues as customers shift
usage from peak to off-peak pricing periods. The reduction in revenues from the peakperiod usage reductions exceeds the increase in revenues from the increase in off-peak
period usage increases, reducing total utility revenues. However, the magnitude of the
revenue attrition due to these rates is small, ranging from 0.3 to 0.8 percent of current
revenue.
Table 6.3: Revenue Attrition Due to Customer Demand Response, Expected Elasticity
Rate Structure
Utility
Midwest
KCP&L
Westar
SFV
IBR
TOU
2.2%
1.6%
2.9%
-3.7%
-3.2%
-4.6%
n/a
-0.4%
-0.4%
Day-Type
TOU
n/a
-0.5%
-0.5%
Table 6.4: Revenue Attrition Due to Customer Demand Response, High Elasticity
Rate Structure
Utility
Midwest
KCP&L
Westar
SFV
IBR
TOU
4.4%
3.3%
5.9%
-7.3%
-6.4%
-7.0%
n/a
-0.6%
-0.6%
Day-Type
TOU
n/a
-0.8%
-0.8%
Table 6.5: Revenue Attrition Due to Customer Demand Response, Low Elasticity
Rate Structure
Utility
Midwest
KCP&L
Westar
SFV
IBR
TOU
1.1%
0.8%
1.5%
-1.9%
-1.6%
-3.3%
n/a
-0.2%
-0.2%
Day-Type
TOU
n/a
-0.3%
-0.3%
In summary, the results show the importance of accounting for revenue attrition due to
customer demand response for SFV and IBR rates, but indicate that the issue is not
significant for the TOU and day-type TOU rates.
40
CA Energy Consulting
7. Summary and Conclusions
This report analyzed the effects associated with changing residential rate structures for
residential customers at KCP&L, Westar, and Midwest Energy. The study includes five
rate structures:
• Flat rates, which charge the same rate in all hours;
• Straight-fixed variable (SFV) rates, in which the customer charge is increased to
recover all fixed costs;
• Inclining-block rates (IBR), in which the rate increases with the level of usage;
• Time-of-use (TOU) rates, in which the rate varies by time of day; and
• Day-type TOU rates, in which the rate varies by time of day and with system
conditions.
For each rate structure and utility, we designed a rate that is revenue neutral based on the
representative samples of customers available to us. We then calculated customer-level bill
impacts and presented figures that illustrate the relationship between the percentage bill
impact and the customer usage level (or size). We found a strong relationship between
customer size and bill impacts for some of the rate structures:
• SFV rates increase bills for low-use customers and decrease bills for high-use
customers;
• IBR rates tend to decrease bills for low-use customers and increase bills for highuse customers; and
• A combination of SFV and IBR produces bill impacts that more closely resemble
the SFV bill impacts.
TOU and day-type TOU bill impacts are not strongly related to customer size. Rather, they
tend to benefit customers with relatively less usage during peak hours.
The range of bill impacts (highest to lowest) was significantly higher for SFV and IBR
than for TOU and day-type TOU rates. For example, the percentage bill impact on
KCP&L's SFV rate ranged from -7.7 percent to 50.1 percent; while the bill impacts on its
TOU rate ranged from -3.9 percent to 13.3 percent.
In addition to analyzing bill impacts, we conducted a high-level simulation of the overall
usage and load impacts that may be expected to occur for each rate structure. Because SFV
and IBR change the rate in all hours, they are modeled as affecting overall load changes,
with SFV tending to increase usage 20 and IBR tending to decrease usage. Because TOU
and day-type TOU rates change by time of day, we modeled the effects of these rates as
shifts of usage from peak to off-peak periods.
Finally, the report examined the potential for utility revenue attrition (recovering less
revenue than forecast) due to customer self selection and demand response. That is, when
the utility sets the rates for an optional pricing program, it does not know which customers
20
The load change from SFV does not include any effects associated with increases in utility-assisted
conservation efforts. A primary goal of SFV is to remove the utility's disincentive to promote conservation.
41
CA Energy Consulting
will select the rate, or how the customers who select the rate will modify their load profiles
in response to the new price signals. Our analysis provided an indication of the scale of
this potential problem by assuming that customers select the rate that provides them with
the lowest bill (customer self selection); and by simulating customer demand response
using a range of price responsiveness parameters (i.e., price elasticities). The results
indicated that both types of revenue attrition (i.e., due to customer self selection and
demand response) are more pronounced for SVF and IBR than they are for TOU and daytype TOU.
42
CA Energy Consulting